yama@criepi.denken.or.jp

Identification of Individual Load Self-disconnection Following a Voltage Sag

K. YAMASHITA*, H. KOBAYASHI, and Y. KITAUCHI

Central Research Institute of Electric Power Industry

Japan

SUMMARY

Voltage sag is one of the main causes of load self-disconnection, a large amount of which could

adversely affect the short-term voltage stability and transient stability of power systems. In 1987, the

Electric Technology Research Association (ETRA) set up a national technical committee in Japan to

clarify the recent characteristics of voltage sag and load self-disconnection following a three-phase

fault, and to identify countermeasures against voltage sag. Although the technical committee compiled

an excellent report which is still widely referred to, the results are becoming obsolete.

Due to the increasing percentage of inverter-based loads, there is much interest in the influence of

voltage sag on the dynamic behavior of the rapid rise in post-fault load voltage for identifying

technical issues concerning voltage phenomena. Therefore, in 2010 ETRA set up a national technical

committee for reviewing past results. One important task of the committee was to update the load self-

disconnection characteristics using modern loads. ETRA requested the Central Research Institute of

Electric Power Industry (CRIEPI) to prepare for more than 40 kinds of load which are widely used in

Japan with the aid of committee members in order to obtain load self-disconnection characteristics

through field tests at the Akagi Testing Center of CRIEPI.

A designated voltage dip (10% to 80%, in increments of 10%) and designated duration of voltage sag

(2 ms to 500 ms) were applied to loads by a back-to-back (BTB) controller. Some of the loads started

to be disconnected from the power system when the voltage sag exceeded 20% and no more loads

could be disconnected when the voltage sag exceeded 70%. Overall, inverter-based air conditioners

and refrigerators tended to be easily disconnected from the power system compared with non-inverter-

based air conditioners and refrigerators. On the other hand, non-inverter-based lamps tended to be

easily disconnected from the power system compared with inverter-based lamps.

It is known that photovoltaic power generation systems (PV) are often disconnected from the power

system due to the voltage sag caused by their own protection devices. As the use of PV spreads, this

self-disconnection of PV could affect the transmission line system as well as distribution system.

Because a voltage sag causes self-disconnection of both PV and load at the same instant, it is

important to understand load self-disconnection characteristics more precisely for a dynamic security

assessment assuming widespread use of PV in the power system. The results will therefore be useful

for identifying technical issues concerning various phenomena in the power system in the future.

KEYWORDS

Load self-disconnection, voltage sag, inverter-based load, power system

Oct.26-28, 2011, Thailand OP-08 CIGRE-AORC 2011

www.cigre-aorc.com

1

1. INTRODUCTION

Voltage sag is one of the main causes of load self-disconnection, a large amount of which could

adversely affect the short-term voltage stability and transient stability of power systems. In 1987, the

Electric Technology Research Association (ETRA) set up a national technical committee in Japan to

clarify the recent characteristics of voltage sag and load self-disconnection following a three-phase

fault, and to identify countermeasures against voltage sag [1]. Although the technical committee

compiled an excellent report which is still widely referred to, the results are becoming obsolete.

Due to the increasing percentage of inverter-based loads, there is much interest in the influence of

voltage sag on the dynamic behavior of the rapid rise in post-fault load voltage for identifying

technical issues concerning voltage phenomena. Therefore, in 2010 ETRA set up a national technical

committee for reviewing past results [2]. One important task of the committee was to update the load

self-disconnection characteristics using modern loads. ETRA requested the Central Research Institute

of Electric Power Industry (CRIEPI) to prepare for more than 40 kinds of load which are widely used

in Japan with the aid of committee members in order to obtain load self-disconnection characteristics

through field tests at the Akagi Testing Center of CRIEPI. Not only load self-disconnection

characteristics but also their recovery characteristics were investigated through the field tests.

2. SELECTION OF LOADS UNDER TEST

Loads were selected based on the four types shown in Table 1. The non-inverter-based loads that were

selected in 1987 were firstly selected as the loads under test (LUTs). Inverter-based loads that did not

exist and were not used in 1987 were also selected as LUTs. The LUTs are shown in Table 2. LUTs

were classified into two groups: 1) LUTs which were expected to take only a few minutes to recover

following a three-phase fault, and 2) LUTs which were expected to take more than a few minutes to

recover.

Table 1 Classification of Loads Under Test Type Example

Information, communication

and electronics equipment PC, Ethernet hub, DVD recorder, television

Electrically-powered equipment Electromagnetic switch, protective relay, signal relay, timer

Lamps Fluorescent lamp, light emitting diode (LED) lamp, discharge lamp

Thermoelectric equipment Induction heating equipment, air-conditioner, refrigerator

Table 2 List of Loads Under Test Type of Loads Product and Description

(1) Computer for personal use DELL: OPTIPLEX GX60 115/230 2/1A

(2) Computer for industrial use HITACHI: HF-W6500

(3) Ethernet hub BUFFALO: LSW2-GT-16NSRR 100V 16.5W

(4) Digital TV TOSHIBA: 37Z7000 100V 239W made in 2009

(5) Digital TV SHARP: LC-20D30 100V 72W made in 2008

(6) DVD recorder SONY: RDR-VH95 100V 38W made in 2006

(7) Control sequencer for industrial use OMRON: SYSMAC CJ1M CPU11 (SCU21-V1, OC211, PA202) 100-240V

(8) Electromagnetic switch PANASONIC: BMF6-100 AC100V, Three-phase 200V 9kW

(9) Electromagnetic switch Manufacturer A: 220-240VAC, Three-phase 220V-150A

(10) Electromagnetic switch Manufacturer A: 100-240VAC, Three-phase 220V-32A

(11) Numerical relay TOSHIBA: Earth-fault Overvoltage Relay NVG21S-01A51 made in 2009

(12) Miniature relay / Signal relay OMRON: MY4ZN-CR AC100/110V

(13) Miniature relay / Signal relay OMRON: MY2N-Y AC200/220V

(14) Miniature relay / Signal relay OMRON: MM2XP AC110V

(15) Miniature relay / Signal relay OMRON: MY4 AC100/110V

(16) Timer OMRON: H3CR-A8 AC100-240V

(17) Fluorescent lamp (inverter-based) TOSHIBA: FSH91353R 100V 110W

(18) High-pressure mercury lamp

(Non-inverter-based) IWASAKI Electric: HF100X 100V 100W

(19) High-pressure discharge lamp IWASAKI Electric: HRF200X 100V 360W

2

(Non-inverter-based)

(20) High-pressure discharge lamp

(Non-inverter-based) IWASAKI Electric: MT1500B-D/BH 242V 1500W

(21) High-pressure discharge lamp

(Inverter-based) TOSHIBA: 200V MF400EB-J2/BU-P 400W

(22) Air conditioner (Inverter-based) TOSHIBA: RAS-402PADR 200V 900W (for cooling), 950W (for heating),

made in 2009

(23) Air conditioner (Non-inverter-based) HITACHI: RAC-22HSFW 100V 870W (for cooling), 860W (for heating),

made in 1997

(24) Air conditioner for industrial use DAIKIN: RYP140AA, Three-phase, 200V 5kW, made in 2009

(25) Refrigerator (Non-inverter-based) SHARP: SJ-17J-H 165L 100V 140W, made in 2005

(26) Refrigerator (Inverter-based) SHARP: SJ-W42DE-H 415L 100V 119W, made in 2002

(27) Refrigerator (Non-inverter-based) HITACHI: R-37V7 370L 100V 150W, made in 1996

(28) Induction heating cooker TOSHIBA: BHP-M46DR, Single-phase 200V 5000W

(29) Induction heating cooker TAKES GROUP: PLM-29700, 100V, 1300W, made in 2006

(30) Electrical hot-water supply system Manufacturer A: Single-phase 200V, 0.915kW, made in 2007

3. TESTING PROCEDURE

LUTs in group 1 or group 2 were connected to a specified load bus (Fig. 1). A designated voltage dip

(10% to 80%, in 10% increments) and designated voltage sag duration (2 ms, 5 ms, 10 ms, 15 ms, 20

ms, 40 ms, 80 ms, 120 ms, 200 ms, 300 ms and 500 ms) were applied to loads via a hybrid simulator

using a back-to-back (BTB) controller (Fig. 2).

Commercial Power System

6.6kV ExperimentalDistribution Line: 1km 1600kVA

BTB

Measurement

Device

V V

I

I

2000kVA

66kV 6.6kV

Single Phase100V/200V

LA

N

Experimental Substation

Experimental Site

Load

1

Load

2

Load

3

Load

4

Load

N

Load BusVoltage SagGenerator

Figure 1 Outline of voltage sag test circuit

Analog SimulatorDigital Simulator

Voltage Control D/A

A/D

Step-up Transformer

DC Voltage Control

Filter

± 10V

± 10V

CPUCPU

CPU CPU

3.3kV

PWM

PWM

ACPower Supply

Interfacing Circuit

Figure 2 Scheme of hybrid simulator using back-to-back controller

4. RESULTS OF THE TEST

As shown in Fig. 3, some of the loads start to be disconnected from the power system when the

voltage sag exceeds 20%. Figure 3 also reveals that no more loads are disconnected when the voltage

sag exceeds 70%. Overall, inverter-based air conditioners and refrigerators tend to be easily

disconnected from the power system compared with non-inverter-based air conditioners and

refrigerators (Fig. 4). On the other hand, non-inverter-based lamps tend to be easily disconnected from

the power system compared with inverter-based lamps (Fig. 5). Most of the miniature relays / signal

3

relays were disconnected by even quite a short voltage sag of less than 20 ms when the sag exceeded

40%. M

agn

itu

de

Sag

to

[%

]

90

80

70

60

50

40

30

20

100.001 0.01 0.1 1

Sag Duration [s]

(24) (30)(23)

(2)(22)(8)

(16)

(26)

(15) (10)

(5) (11)

(27）(1)

(28)

(6)

(4)

(9)(13)

(7)

(20)(19)

(3)(17)

(25)

(14)(12)

(21)

(12) (7)

(17)

(18)

(18)(30)

(18)

(4)(28)

(14)

(29)

Notes:

1: Figures in parentheses in Fig. 1 correspond to those in Table 1.

2: Figures with underlines mean that the LUTs show two different boundaries.

3: Solid lines in the figure denote boundaries at which each LUT runs without self-disconnection.

Figure 3 Individual load self-disconnection characteristics following a voltage sag

2 5 10 15 20 40 80 120 200 300 400 50010 - - - - - - - - - - - -

20 - - - - - - - - - - - -

30 - - - - - - - - - - - -

40 - - - - - - - - - × × ×

50 - - - - - - - - × × × ×

60 - - - - - - - - × × - -

70 - - ▲ - - - - - × × (×) (×)

80 - - ▲ - - - - - × (×) (×) ×

Vo

ltag

e S

ag

[%

]

Fault Duration Time [ms]

2 5 10 15 20 40 80 120 200 300 400 500

10 - - - - - - - - - - - -

20 - - - - - - - - - - - -

30 - - - - - - - - - - - -

40 - - - - - - - - - - - -

50 - - - - - - - - - - - -

60 - - - - - - - - - - - -

70 - - ▲ - - - - - - - - -

80 - - ▲ - - - - - - - × ×

Vo

ltag

e S

ag

[%

]

Fault Duration Time [ms]

▲: Unadministered test ▲: Unadministered test

x: Restart after 5–6 minutes stoppage x: Restart after 1 second stoppage

(x): Restart after a few seconds stoppage

Figure 4 Example of load self-disconnection characteristics following a voltage sag

(Left: Inverter-based refrigerator (26), Right: Non-inverter-based refrigerator (25))

2 5 10 15 20 40 80 120 200 300 400 500

10 - ▲

△ △ △ △ △ △ △ △ △ △

20 - ▲

△ △ △ △ △ △ △ △ △ △

30 - ▲

△ △ △

× × × × × × ×

40 -

△ △

× × × × × × × × ×

50 -

△ △

× × × × × × × × ×

60 -

△ △

× × × × × × × × ×

70 -

△

▲ × × × × × × × × ×

80 -

△

▲ × × × × × × × × ×

Vo

ltag

e S

ag

[%

]

Fault Duration Time [ms]

2 5 10 15 20 40 80 120 200 300 400 500

10 - - - - - - - - - - - -

20 - - - - - - - - - - - -

30 - - - - - - - - -

△ △ △

40 - - - - - -

△ △ △ △ △ △

50 - - - - - -

△ △ △ △ △ △

60 - - - - - -

△ △ △ △ △ △

70 - - ▲ -

△ △ △ △ △ △ △ △

80 - - ▲ -

△ △ △ △ △ △

× ×

Vo

ltag

e S

ag

[%

]

Fault Duration Time [ms]

▲: Unadministered test ▲: Unadministered test

x: Restart after 4 minutes stoppage x: Restart after 5–6 minutes stoppage

∆: Instantaneous extinction ∆: Instantaneous extinction

Figure 5 Example of load self-disconnection characteristics following a voltage sag

(Left: Non-inverter-based high-pressure mercury lamp (19),

Right: Inverter-based high-pressure mercury lamp (21))

4

Because the field test was performed in December 2010, the air-conditioners operated in heating mode.

The load self-disconnection characteristics of air-conditioners operated in cooling mode are shown in

reference [3]. The compressors of both inverter-based and non-inverter-based air-conditioners stopped

for a few minutes (between 2 and 5 minutes) due to a severe voltage sag and automatically restarted

(Fig. 6). While the compressors were stopped, the indoor fans of both inverter-based and non-inverter-

based air-conditioners continued to operate. The compressors of refrigerators stopped for about 5-6

minutes due to a severe voltage sag and automatically restarted (Fig. 7).

Note that the non-inverter-based air-conditioners and refrigerators increased their apparent power

output for 2 to 20 seconds after the voltage sag recovered (Fig. 6). The post-fault behavior of the non-

inverter-based air-conditioner is considered to be due to the dynamic behavior of a conventional

induction motor which often causes a slow voltage recovery.

With regard to the load recovery characteristics, it took approximately 10 minutes for the inverter-

based air-conditioner to recover its active power output, while it took less than 2 minutes for the non-

inverter-based air-conditioner to do so (Fig. 6). Moreover, it took just several seconds for the inverter-

based and non-inverter-based refrigerators to recover their active power output.

-200

-1000

100

200

VU

-N[V

]

2520151050

Time [s]

-20

0

20

I [A

]

1000

500

0

-500

P [

W]

200

100

0Q [

var

]

-200

-1000

100

200V

U-N

[V]

2520151050

Time [s]

-100

0

100

I [A

]

2000

1000

0

-1000

P [

W]

-1000

-500

0

500

Q [

var

]

Sudden Increase of Current Before Stoppage

800

400

0

P [

W]

8006004002000-200

Time [s]

200

100

0

-100

Q [

var

]

CompressorStoppage

During Restart

1500

1000

500

0

P [

W]

2001000-100

Time [s]

200

0

-200

Q [

var

]

CompressorStoppage During Restart

Figure 6 Example of load self-disconnection characteristics following a voltage sag

(Left: Inverter-based air conditioner under 40% voltage sag and 500 ms fault duration (22),

Right: Non-inverter-based air conditioner under 80% voltage sag and 200 ms fault duration (23))

Digital televisions showed different load recovery characteristics. When digital televisions are turned

on by pressing the power button on the remote control, the name of the manufacturer is often

displayed for several seconds before a TV program appears. This time period can be considered as

preparation for the starting process and a small amount of active power is consumed. When a digital

television stops due to a severe voltage sag, the status becomes the same as turning off the television

by the remote control. In other words, the active power output of the television increases in two

different stepwise changes until it completely recovers.

5

-200

-1000

100

200V

U-N

[V]

2520151050

Time [s]

-20-10

01020

I [A

]

400

200

0

-200

P [

W]

100

50

0

Q [

var

]

-200

-1000

100

200

VU

-N[V

]

2520151050

Time [s]

-20-10

01020

I [A

]

200

100

0

-100

P [

W]

100

50

0

-50

Q [

var

] Preparation of Restart

Restart

200

100

0

P [

W]

6004002000-200

Time [s]

302010

0

Q [

var

]

CompressorStoppage

Restart

2 5 10 15 20 40 80 120 200 300 400 50010 - - - - - - - - - - - -

20 - - - - - - - - - - - -

30 - - - - - - - - - - - -

40 - - - - - - - - - - - -

50 - - - - - - - - - - - -

60 - - - - - - - - - - - -

70 - - ▲ - - - - - × × × ×

80 - - ▲ - - - × × × × × ×

Vo

ltag

e S

ag

[%

]

Fault Duration Time [ms]

▲: Unadministered test

x: Restart immediately after voltage recovery

Figure 7 Example of load self-disconnection characteristics following a voltage sag

(Left: Inverter-based refrigerator under 70% voltage sag and 500 ms fault duration (26),

Right: Digital television under 80% voltage sag and 500 ms fault duration (5))

Electromagnetic switches are widely used for conventional induction motors. In the tests, the

electromagnetic switches were once disconnected (the contactors opened) due to the severe voltage

sag, then were connected again immediately after the voltage recovered. The load self-disconnection

characteristics of electromagnetic switches are similar to those of conventional induction motors. Note

that once an electromagnetic switch equipped in an induction motor is disconnected, the motor does

not automatically restart even though the electromagnetic switch is reconnected immediately after the

voltage recovers.

2 5 10 15 20 40 80 120 200 300 400 50010 - - - - - - - - - - - -

20 - - - - - - - - - - - -

30 - - - - - - - - - - - -

40 - - - - - - - - - - - -

50 - - - - - - - × × × × ×

60 - - - - - - × × × × × ×

70 - - ▲ - - - × × × × × ×

80 - - ▲ - - × × × × × × ×

Fault Duration Time [ms]

Vo

ltag

e S

ag

[%

]

2 5 10 15 20 40 80 120 200 300 400 50010 - - - - - - - - - - - -

20 - - - - - - - - - - - -

30 - - - - - - - - - - - -

40 - - - - - - - - - - - -

50 - - - - - - - - - - - -

60 - - - - × × × × × × × ×

70 - - ▲ - × × × × × × × ×

80 - - ▲ - × × × × × × × ×

Fault Duration Time [ms]

Vo

ltag

e S

ag

[%

]

▲: Unadministered test ▲: Unadministered test

x: Reconnection immediately after voltage recovery x: Reconnection immediately after voltage recovery

Figure 8 Example of load self-disconnection characteristics following a voltage sag

(Left: Electromagnetic switch (9), Right: Electromagnetic switch (10))

5. CONCLUSIONS

It is known that PV systems are often disconnected from the power system due to the voltage sag

caused by their own protection devices. As the use of PV spreads, this self-disconnection of PV could

6

affect the transmission line system as well as distribution system. Because a voltage sag causes self-

disconnection of both PV and load at the same instant, it is important to understand load self-

disconnection characteristics more precisely for a dynamic security assessment assuming widespread

use of PV.

Experiments at CRIEPI in 2010 clarified that some of the loads start to be disconnected from the

power system when the voltage sag exceeds 20% and that no more loads can be disconnected when

the voltage sag exceeds 70%. Overall, inverter-based air conditioners and refrigerators tend to be

easily disconnected from the power system compared with non-inverter-based air conditioners and

refrigerators. On the other hand, non-inverter-based lamps tend to be easily disconnected from the

power system compared with inverter-based lamps. It is concluded that if the usage of air-conditioners

increases, the amount of self-disconnected loads due to a severe voltage sag will increase.

Thus, Fig. 3 is useful for identifying technical issues concerning various power system phenomena in

power systems in the future, and will also contribute to the ongoing CIGRE WG C4.605 “Modelling

and aggregation of loads in flexible power networks.”

REFERENCES

[1] Electric Technology Research Association: “Countermeasure for voltage dips in power

systems,” Vol. 46, No. 3, 1990.

[2] Electric Technology Research Association: “Countermeasure technology for voltage dips in

power systems” Vol. 67, No. 2, 2011.

[3] K. Yamashita and O. Sakamoto: “A Study on Dynamic Behavior of Load Supply System

including Synchronous Generators with and without Load Drop,” Proceedings of IEEE 2010

PES General Meeting, 2010.

Short Biography of Main Author

Koji Yamashita received his B.S. and M.S. degrees from Waseda University, Tokyo,

Japan, in 1993 and 1995, respectively. Since 1995, he has been with the Department of

Power Systems at Central Research Institute of Electric Power Industry in Tokyo,

Japan. He is a regular member of CIGRE C4.605 WG.

1

Id tifi ti f I di id l L dId tifi ti f I di id l L dIdentification of Individual Load Identification of Individual Load SelfSelf disconnectiondisconnectionSelfSelf--disconnection disconnection

Following a Voltage SagFollowing a Voltage Sagg g gg g g

Koji Yamashita (CRIEPI)Koji Yamashita (CRIEPI)H Kobayashi (CRIEPI)Y Kitauchi (CRIEPI)

26th-27th of October 2011, Chiang Mai, Thailand

Introduction (Cont’d) Voltage sag is one of the main causes of load self-

disconnection, a large amount of which could adverselydisconnection, a large amount of which could adversely affect the short-term voltage stability and transient stability of power systems.of power systems.

In 1987, the Electric Technology Research Association (ETRA) set up a national technical committee in Japan to(ETRA) set up a national technical committee in Japan to clarify the recent characteristics of voltage sag and load self-disconnection following a three-phase fault and toself-disconnection following a three-phase fault, and to identify countermeasures against voltage sag.

Alth h th t h i l itt il d ll t Although the technical committee compiled an excellent report which is still widely referred to, the results are becoming obsoletebecoming obsolete.

Introduction Due to the increasing percentage of inverter-based loads,

there is much interest in the influence of voltage sag on thethere is much interest in the influence of voltage sag on the dynamic behavior of the rapid rise in post-fault load voltage for identifying technical issues concerning voltagevoltage for identifying technical issues concerning voltage phenomena.

In 2010 ETRA set up a national technical committee for In 2010 ETRA set up a national technical committee for reviewing past results and for updating the load self-disconnection characteristics using modern loadsdisconnection characteristics using modern loads.

Not only load self-disconnection characteristics but also th i h t i ti i ti t d th h ththeir recovery characteristics were investigated through the field tests

Classification of Loads Under Test (LUTs)4

Type ExampleInformation,Information,

communicationand electronics PC, Ethernet hub, DVD recorder, television

equipmentElectrically-powered

equipmentElectromagnetic switch, protective relay, signal

relay timerequipment relay, timer

Lamps Fluorescent lamp, light emitting diode (LED) lamp, discharge lampp, g p

Thermoelectric equipment

Induction heating equipment, air-conditioner, refrigerator

The non-inverter-based loads that were selected in 1987 were firstly selected as the loads under test (LUTs). Inverter-based loads that did

not exist and were not used in 1987 were also selected as LUTs.

List of Loads Under Test (Con’d)5

Type of Loads Product and Description(1) Computer for personal use DELL: OPTIPLEX GX60 115/230 2/1A(2) C t f i d t i l HF W6500(2) Computer for industrial use : HF-W6500(3) Ethernet hub : LSW2-GT-16NSRR 100V 16.5W(4) Digital TV TOSHIBA: 37Z7000 100V 239W made in 2009(5) Digital TV SHARP: LC 20D30 100V 72W made in 2008(5) Digital TV SHARP: LC-20D30 100V 72W made in 2008 (6) DVD recorder SONY: RDR-VH95 100V 38W made in 2006(7) Control sequencer for industrial use

OMRON: SYSMAC CJ1M CPU11 (SCU21-V1, OC211, PA202) 100-240Vuse PA202) 100 240V

(8) Electromagnetic switch PANASONIC: BMF6-100 AC100V, Three-phase 200V 9kW(9) Electromagnetic switch Manufacturer A: 220-240VAC, Three-phase 220V-150A(10) Electromagnetic switch Manufacturer A: 100-240VAC, Three-phase 220V-32A(10) Electromagnetic switch Manufacturer A: 100 240VAC, Three phase 220V 32A

(11) Numerical relay TOSHIBA: Earth-fault Overvoltage Relay NVG21S-01A51 made in 2009

(12) Miniature relay / Signal relay OMRON: MY4ZN-CR AC100/110V( ) y g y(13) Miniature relay / Signal relay OMRON: MY2N-Y AC200/220V(14) Miniature relay / Signal relay OMRON: MM2XP AC110V(15) Miniature relay / Signal relay OMRON: MY4 AC100/110V(16) Timer OMRON: H3CR-A8 AC100-240V(17) Fluorescent lamp (inverter-based) TOSHIBA: FSH91353R 100V 110W

6Type of Loads Product and Description

(18) High-pressure mercury lamp(Non inverter based) IWASAKI Electric: HF100X 100V 100W(Non-inverter-based)(19) High-pressure discharge lamp(Non-inverter-based) IWASAKI Electric: HRF200X 100V 360W

(20) High pressure discharge lamp(20) High-pressure discharge lamp(Non-inverter-based) IWASAKI Electric: MT1500B-D/BH 242V 1500W

(21) High-pressure discharge lamp(Inverter-based) TOSHIBA: 200V MF400EB-J2/BU-P 400W(Inverter based)

(22) Air conditioner (Inverter-based) TOSHIBA: RAS-402PADR 200V 900W (for cooling), 950W (for heating), made in 2009

(23) Air conditioner (Non-inverter- : RAC-22HSFW 100V 870W (for cooling), 860W (for ( 3) co d t o e (No ve tebased)

: C S W 00V 870W ( o coo g), 860W ( oheating), made in 1997

(24) Air conditioner for industrial use

DAIKIN: RYP140AA, Three-phase, 200V 5kW, made in 2009

(25) Refrigerator (Non-inverter-based) SHARP: SJ-17J-H 165L 100V 140W, made in 2005

(26) Refrigerator (Inverter-based) SHARP: SJ-W42DE-H 415L 100V 119W, made in 2002(27) Refrigerator (Non-inverter-based) : R-37V7 370L 100V 150W, made in 1996

(28) Induction heating cooker TOSHIBA: BHP-M46DR, Single-phase 200V 5000W(29) Induction heating cooker TAKES GROUP: PLM-29700, 100V, 1300W, made in 2006(30) Electrical hot-water supply system Manufacturer A: Single-phase 200V, 0.915kW, made in 2007

Outline of voltage sag test circuit7

Note: LUTs in group 1 or group 2 were connected to a specified load bus

ETRA requested the Central Research Institute of Electric Power

Note: LUTs in group 1 or group 2 were connected to a specified load bus

ETRA requested the Central Research Institute of Electric Power Industry (CRIEPI) to prepare for more than 40 kinds of load which are widely used in Japan with the aid of committee members in order to y pobtain load self-disconnection characteristics through field tests at the Akagi Testing Center of CRIEPI.

Scheme of hybrid simulator using back-to-back (BTB) controller

8

using back to back (BTB) controller

A designated voltage dip (10% to 80%, in 10% increments) and designated voltage sag duration (2 ms, 5 ms, 10 ms, 15 ms, 20 ms, 40 g g g (ms, 80 ms, 120 ms, 200 ms, 300 ms and 500 ms) were applied to loads via a hybrid simulator using a back-to-back (BTB) controller

Fig.3 Individual load self-disconnection characteristics following a voltage sag

9

characteristics following a voltage sag

(24) (30)(23)

(2)(22)(8) (26)

(20)(19)

(12) (7) (18)(30)

(18)

(14) (2)(22)(8)

(16)

(26)

(4)

(9)(13)(14)(12)

(12) (7)

(17)

(18)

(30)(14)

(15) (10)

(5) (11)

(28)

(6)

(4)

(7) (25)

(18)

(29)

(27）(1)(3)(17) (21)(4)(28)

Some of the loads start to be disconnected from the power system when So e o t e oads sta t to be d sco ected o t e powe syste w ethe voltage sag exceeds 20%. Figure 3 also reveals that no more loads are disconnected when the voltage sag exceeds 70%.

Inverter-based refrigerator (26) load self-disconnection characteristics following a voltage sag

10

characteristics following a voltage sag

2 5 10 15 20 40 80 120 200 300 400 500Fault Duration Time [ms]

2 5 10 15 20 40 80 120 200 300 400 50010 - - - - - - - - - - - -

20 - - - - - - - - - - - -%]

2030 - - - - - - - - - - - -

40 - - - - - - - - - × × ×Sag

[%

50 - - - - - - - - × × × ×

60 - - - - - - - - × × - -

70 ▲ × × (×) (×)Vol

tage

70 - - ▲ - - - - - × × (×) (×)

80 - - ▲ - - - - - × (×) (×) ×

V

▲: Unadministered test▲: Unadministered testx: Restart after 5–6 minutes stoppage (x): Restart after a few seconds stoppage

Inverter-based ACs and refrigerators tend to be easily disconnected from the power system compared with non-inverter-based ACs and fridges

( ) pp g

Non-inverter-based refrigerator (25) load self-disconnection characteristics following a voltage sag

11

disconnection characteristics following a voltage sag

2 5 10 15 20 40 80 120 200 300 400 500Fault Duration Time [ms]

2 5 10 15 20 40 80 120 200 300 400 50010 - - - - - - - - - - - -

20 - - - - - - - - - - - -%] 20

30 - - - - - - - - - - - -

40 - - - - - - - - - - - -

e Sa

g [%

50 - - - - - - - - - - - -

60 - - - - - - - - - - - -

70 ▲Vol

tage

▲: Unadministered test

70 - - ▲ - - - - - - - - -

80 - - ▲ - - - - - - - × ×

V

▲: Unadministered testx: Restart after 1 second stoppage

Inverter-based ACs and refrigerators tend to be easily disconnected from the power system compared with non-inverter-based Acs and fridges

Non-inverter-based high-pressure mercury lamp (19) load self-disconnection characteristics following a voltage sag

12

self-disconnection characteristics following a voltage sag

2 5 10 15 20 40 80 120 200 300 400 500Fault Duration Time [ms]

2 5 10 15 20 40 80 120 200 300 400 50010 - ▲

20 - ▲

[%]

30 - ▲ × × × × × × ×

40 - × × × × × × × × ×

50 × × × × × × × × ×e Sa

g [

50 - × × × × × × × × ×

60 - × × × × × × × × ×

70 - ▲ × × × × × × × × ×Vol

tag

▲: Unadministered test

7080 - ▲ × × × × × × × × ×

V

▲: Unadministered testx: Restart after 4 minutes stoppage∆: Instantaneous extinction

Non-inverter-based lamps tend to be easily disconnected from the power system compared with inverter-based lamps

Inverter-based high-pressure mercury lamp (19) load self-disconnection characteristics following a voltage sag

13

disconnection characteristics following a voltage sag

2 5 10 15 20 40 80 120 200 300 400 500Fault Duration Time [ms]

2 5 10 15 20 40 80 120 200 300 400 50010 - - - - - - - - - - - -

20 - - - - - - - - - - - -%]

30 - - - - - - - - -

40 - - - - - -

e Sa

g [%

50 - - - - - -

60 - - - - - -

70 - - ▲ -Vol

tage

▲: Unadministered test

70 - - ▲ -

80 - - ▲ - × ×

V

▲: Unadministered testx: Restart after 5–6 minutes stoppage∆: Instantaneous extinction

Non-inverter-based lamps tend to be easily disconnected from the power system compared with inverter-based lamps

Inverter-based AC (22) load self-disconnection characteristics following a voltage sag

14

characteristics following a voltage sag

0100200

[V]

-200-100

0

VU

-N

20

40% voltage sag 500 ms fault duration

-200

20

I [A

]

10001000500

0-500

P [W

]

200100

0Q [v

ar]

2520151050Time [s]

0

The compressors of both inverter-based and non-inverter-based air-conditioners stopped for a few minutes and automatically restarted

Non-inverter-based AC (23) load self-disconnection characteristics following a voltage sag

15

characteristics following a voltage sag

100200

[V]

80% voltage sag -200-100

0

VU

-N[

100 Sudden Increase of Current Before Stoppage % g g200 ms fault duration

-100

0

00

I [A

]

2000 The non-inverter-based 20001000

0-1000

P [W

] air-conditioners and refrigerators increased

1000

1000-500

0500

Q [v

ar] their apparent power

output for 2 to 20 seconds after the

2520151050Time [s]

-1000 seconds after the voltage sag recovered

[ ]

The compressors of both inverter-based and non-inverter-based air-conditioners stopped for a few minutes and automatically restarted

Inverter-based AC (22) load self-disconnection characteristics following a voltage sag

16

characteristics following a voltage sag800]

CompressorStoppage

During Restart

400

0

P [W

Stoppage

0200100

0[var

]

8006004002000-200

0-100

Q

40% voltage sag

8006004002000200Time [s]

40% voltage sag 500 ms fault duration

It took approximately 10 minutes for the inverter-based air-conditioner to recover its active power output

Non-inverter-based AC (23) load self-disconnection characteristics following a voltage sag

17

characteristics following a voltage sag1500

] 1000500

0

P [W

]

CompressorStoppage During Restart0

2000[v

ar]

2001000100

0-200Q

80% l

2001000-100Time [s]

80% voltage sag 200 ms fault duration

It took less than 2 minutes for the non-inverter-based air-conditioner to recover its output

Inverter-based refrigerator (26) load self-disconnection characteristics following a voltage sag

18

characteristics following a voltage sag

100200

[V]

-200-100

0

VU

-N[

20

70% voltage sag -20-10

010

0

I [A

]

400 500 ms fault duration 400200

0-200

P [W

]

200100

50

0Q [v

ar]

2520151050Time [s]

0

The compressors of refrigerators stopped due to a severe voltage sag.

[ ]

Inverter-based refrigerator (26) load self-disconnection characteristics following a voltage sag

19

characteristics following a voltage sag200

] Restart100

0

P [W

]

0302010[v

ar]

CompressorStoppage

6004002000200

100Q

Stoppage

70% voltage sag 500 ms fault duration

6004002000-200Time [s]

The compressors of fridges stopped for about 5-6 minutes due to a severe voltage sag and automatically restarted.

It took just several seconds for the inverter-based and non-inverter-based refrigerators to recover their active power output.

Digital television (5) load self-disconnection characteristics following a voltage sag

20

characteristics following a voltage sag

0100200

N[V

]

-200-100

0

VU

-N

1020

80% voltage sag -20-10

010

I [A

]

200Restart

500 ms fault duration 200100

0-100

P [W

]

10050

050

Q [v

ar] Preparation of Restart

2520151050Time [s]

-50

The active power output of the television increases in two different stepwise changes until it completely recovers

Digital television (5) load self-disconnection characteristics following a voltage sag

21

characteristics following a voltage sag

2 5 10 15 20 40 80 120 200 300 400 500

Fault Duration Time [ms]2 5 10 15 20 40 80 120 200 300 400 500

10 - - - - - - - - - - - -

20 - - - - - - - - - - - -%]

2030 - - - - - - - - - - - -

40 - - - - - - - - - - - -

e Sa

g [%

50 - - - - - - - - - - - -

60 - - - - - - - - - - - -

70 ▲ × × × ×Vol

tage

▲: Unadministered test

70 - - ▲ - - - - - × × × ×

80 - - ▲ - - - × × × × × ×

V

▲: Unadministered testx: Restart immediately after voltage recovery

Digital television tend to be easily disconnected from the power system compared with conventional analog television.

Electromagnetic switch (9) load self-disconnection characteristics following a voltage sag

22

characteristics following a voltage sag

2 5 10 15 20 40 80 120 200 300 400 500

Fault Duration Time [ms]2 5 10 15 20 40 80 120 200 300 400 500

10 - - - - - - - - - - - -

20 - - - - - - - - - - - -%]

2030 - - - - - - - - - - - -

40 - - - - - - - - - - - -

e Sa

g [%

50 - - - - - - - × × × × ×

60 - - - - - - × × × × × ×

70 ▲ × × × × × ×Vol

tage

▲: Unadministered test

70 - - ▲ - - - × × × × × ×

80 - - ▲ - - × × × × × × ×

V

The electromagnetic switches were once disconnected

▲: Unadministered testx: Restart immediately after voltage recovery

The electromagnetic switches were once disconnected due to the severe voltage sag, then were connected again

immediately after the voltage recovered

Electromagnetic switch (10) load self-disconnection characteristics following a voltage sag

23

characteristics following a voltage sag

2 5 10 15 20 40 80 120 200 300 400 500

Fault Duration Time [ms]2 5 10 15 20 40 80 120 200 300 400 500

10 - - - - - - - - - - - -

20 - - - - - - - - - - - -%]

2030 - - - - - - - - - - - -

40 - - - - - - - - - - - -

e Sa

g [%

50 - - - - - - - - - - - -

60 - - - - × × × × × × × ×

70 ▲ × × × × × × × ×Vol

tage

▲: Unadministered test

70 - - ▲ - × × × × × × × ×

80 - - ▲ - × × × × × × × ×

V

▲: Unadministered testx: Restart immediately after voltage recovery

The electromagnetic switches were once disconnected e e ect o ag et c sw tc es we e o ce d sco ecteddue to the severe voltage sag, then were connected again

immediately after the voltage recovered

Conclusion (Cont’d)24

• It is known that PV systems are often disconnected from the power system due to the voltage sag caused by their ownpower system due to the voltage sag caused by their own protection devices.

• As the use of PV spreads this self-disconnection of PV• As the use of PV spreads, this self-disconnection of PV could affect the transmission line system as well as distribution systemdistribution system.

• Because a voltage sag causes self-disconnection of both PV d l d t th i t t it i i t t t d t dand load at the same instant, it is important to understand

load self-disconnection characteristics more precisely for a d i it t i id d fdynamic security assessment assuming widespread use of PV.

Conclusion (Cont’d)25

• Experiments at CRIEPI in 2010 clarified that some of the l d b di d f h hloads start to be disconnected from the power system when the voltage sag exceeds 20% and that no more loads can be di d h h l d 70%disconnected when the voltage sag exceeds 70%.

• Inverter-based air conditioners and refrigerators tend to be easily disconnected from the power system compared with non-inverter-based air conditioners and refrigerators.

• Non-inverter-based lamps tend to be easily disconnected from the power system compared with inverter-based lamps. p y p p

• If the usage of air-conditioners increases, the amount of self-disconnected loads due to a severe voltage sag willself disconnected loads due to a severe voltage sag will increase.

Conclusion26

Fig. 3 is useful for identifying technical issues concerning various power system phenomena in power systems in thevarious power system phenomena in power systems in the future, and will also contribute to the ongoing CIGRE WG C4.605 “Modelling and aggregation of loads in flexible powerC4.605 Modelling and aggregation of loads in flexible power networks.”

27

Thank you for your attentionThank you for your attention.